I just updated to R vs. 4.5.0 on Ubuntu 24.04, requiring pretty much a reinstall of all my packages. I tried to install MetaboAnalystR 4.0.0, but got:
** testing if installed package can be loaded from temporary location
Error: package or namespace load failed for ‘MetaboAnalystR’ in dyn.load(file, DLLpath = DLLpath, …):
unable to load shared object ‘/home/michael/R/x86_64-pc-linux-gnu-library/4.5/00LOCK-MetaboAnalystR/00new/MetaboAnalystR/libs/MetaboAnalystR.so’:
/home/michael/R/x86_64-pc-linux-gnu-library/4.5/00LOCK-MetaboAnalystR/00new/MetaboAnalystR/libs/MetaboAnalystR.so: undefined symbol: match5
Error: loading failed
Execution halted
ERROR: loading failed
I installed it without issue on the previous couple of versions of R (4.4.3 most recently) using:
devtools::install_github("xia-lab/MetaboAnalystR", build = TRUE, build_vignettes = TRUE, build_manual =T)
It doesn’t matter whether I build the vignettes; I still get the error. I will admit I haven’t done a lot to try a track down the source of the error (a project for tomorrow?), but I thought I would pass this on in case it’s an issue with the newest version of R.
Thanks for making and maintaining a great package and a great suite of software.
Were you able to resolve this?
I am also having problems with installation. We run RStudio on a Respublica HPC, and the help team said they tried this both on R 4.2 and 4.4, following the instructions for installation. It says the SSPA package, which is a requirement is not available to bioconductor to either of those versions.
All other Dependencies seem to resolve.
package ’ SSPA’ is not available for Bioconductor version ’ 3.16’
SSPA is no longer required by the latest version of MetaboAnalystR. The message should not occur. You can ignore this message.
I have encount the same issue…
It is not a problem whether ignore it or not, and it is a problem that MetaboanalystR can not be installed successfuly.
Keep in mind that the web-based platform has been designed to eliminate the needs of many such issues. If you choose to use local MetaboAnalystR, make sure you read our post guideline.
It is not convenient for the researcher with more than 100 hundred samples of >150M/files.
On the other way, the figure is not of enough scentific value for publication, and many people wish the Metaboanalyst deployed as local sevice and statistical analysis with R enviroment for figure of high quality…
So, Metaboanalyst on local server or MetaboanalystR is a best choice for the mid-high grade users…
If this is indeed the same issue as the OP, the quick solution is not updating to the most recent version (R 4.5.0) for now, until we find time to look into this.
Regarding the web-based tool deployed locally - it is 10X more challenging and requires a dedicated sys admin. If you really need this support, check out our enterprise solution
This topic was automatically closed after 4 days. New replies are no longer allowed.